2019 Sep-14th

Donald Knuth: “Instead of imagining that our main task is to instruct a computer what to do, let us concentrate rather on explaining to human beings what we want a computer to do.”


  • We need to empower non-statisticians to perform their own data analyses in a reproducible way

  • R is traditionally known for having a steep learning curve, however:

  • Recent developments in the R ecosystem have made it more accessible, allowing us to efficiently teach R to non-statisticians

HealthyR: R for healthcare data analysis

  • Since 2016, primarily in Edinburgh, 2.5 day quick-start course

  • Aimed at clinicians, healthcare data analysts

  • RStudio, tidyverse, relevant datasets

  • Trained over 200 people

Day 1

  • Your first R plot: Quality plots with just 2 lines of code
  • R basics: Read in, modify, create variables
  • Data wrangling: Efficiently summarise data
  • Different types of plots: Bar, box, scatter, line

Days 2-3

  • Working with continuous data (linear regression)
  • Working with categorical data (logistic regression)
  • Time-to-event (survival)
  • Plotting, plotting, plotting

HealthyR Quick-start course set-up

HealthyR Quick-start course set-up



HealthyR Notebooks Objectives

  • Create a lightweight, scalable and open-source resource for clinicians

  • Promote research skills development in low-income settings/those not able to attend an in-person training course

  • Promote reproducible research analyses/automated reports

  • Funded by:

Literate: instructions, R code and output together

Notebooks, laptops, anywhere:



Trialling the new materials

Positives of Notebooks

  • Easier for first-time coders

  • Code is much more clearly annotated

  • Text, code and results are together

  • Data can be published directly from R

More Positives of Notebooks

  • Several formats can be created, thanks Markdown!

  • If the data changes the text updates itself!

  • Carries all benefits of R (free, reproducibile, documentation etc.)

  • Interactive nature is helpful

Some Challenges

  • Installation if not using RStudio cloud, on the normal course we shelter complete beginners from installation frustrations, so not to feed their anxiety

  • Keyboard layouts vary (` and ~ are hard to find!)

  • YAML, chunks, css - yet more things to learn

  • Interacting with colleagues who only use Word (although this could also be a positive/strenght)

Try them yourselves/help your friends to!